Writing · Nº 041065 essays
The writing
Ideas from inside the work. Explore technology, entrepreneurship, and the practice of building things that last.
343 essays found · Page 19 of 29
Clear filtersEngineeringHow to Version and Roll Back an Automation WorkflowWhen an AI workflow change breaks in production, you need to roll it back in seconds. Here is how to version automation workflows so a bad deploy is never permanent.EngineeringHow to Design Pub/Sub Channel GranularityChannel granularity is the biggest pub/sub design decision. Too coarse and you leak data and waste bandwidth; too fine and you drown in subscriptions.EngineeringOffline Evals vs Production Monitoring for AIOffline evals and production monitoring measure different things for an AI feature. You need both. Here is what each catches, what each misses, and how they work together.EngineeringHow to Validate Scraped Data Quality Before It Poisons YouScraped data quality fails silently. A selector drifts and you store nulls for a week. Here is how to validate scraped data at ingestion before it reaches users.EngineeringHow Many WebSocket Connections Fit on One Server?Concurrent WebSocket connections per server are bound by memory and fan-out, not CPU. Learn the capacity math to size real-time infrastructure before you scale out.EngineeringEmail Deliverability Metrics That Actually MatterOpen rate is not deliverability. Here are the email deliverability metrics that actually matter for transactional email and the numbers I watch to keep mail in the inbox.EngineeringShould AI Write Its Own Tests?Should AI write its own tests? Sometimes, but not the ones that verify it. Here is the split I use so agent-written tests help instead of hiding bugs.EngineeringAI Native vs AI Bolted On: How to Tell the DifferenceA decision framework for builders. Tests you can run on your own roadmap to see whether the model is the mechanism of your product or just a feature stapled to an old workflow.EngineeringThe Myth That Public Data Is Free to ScrapePublic does not mean free to take. This breaks down the myth that any public web data is legal to scrape, and what actually determines your real risk.EngineeringSingle-Maintainer Dependency Risk Is the One to WatchOne person maintains the library your product depends on. Here is why single-maintainer dependency risk is worse than an unmaintained one, and what to do.EngineeringOpen Core and BSL License Traps to Watch ForOpen core and BSL license traps can lock you in as hard as any SaaS. How to read software licenses so self-hosting an open tool does not turn into a bill.EngineeringOne Identity Layer Across Every Product You RunA single identity layer across products means one place for auth, sessions, and users. Here is why a shared identity layer beats per-product auth in a portfolio.